In this lesson, you will learn how to subdivide a monolithic landscape actor into smaller, streaming-capable proxies to drastically improve your editor’s responsiveness and reduce runtime memory overhead.
The Problem: The Monolithic Landscape
In massive open-world projects, a single landscape actor containing thousands of components can become a bottleneck. Every time you modify one vertex or adjust a layer, the entire actor undergoes heavy re-calculation. Furthermore, the engine must keep the entire landscape in memory, even if the player is only interacting with a small segment of the world.
The Solution: Landscape Streaming Proxies
Landscape Streaming Proxies allow you to "break off" sections of a primary landscape actor into independent actors. These proxies act as lightweight references to a portion of the original heightmap and material data. By converting your landscape components into proxies, you effectively decouple the landscape into distinct, loadable units that work in tandem with the World Partition system.
To create proxies, select your Landscape actor in the World Outliner, navigate to the Landscape Tool, and find the Manage tab. Look for the Landscape Streaming section. Here, you can select specific components and move them to a new, separate Streaming Proxy actor.
// Conceptually, you are transitioning from this:
LandscapeActor -> [1000 Components]
// To this distributed structure:
MainLandscapeActor -> [Proxy A (Comp 1-100)]
[Proxy B (Comp 101-200)]
[Proxy C (Comp 201-300)]
Performance Benefits
Once subdivided, these proxies respect your grid-based spatial loading cells. Because they are now individual actors, the engine can stream them in and out of memory based on the player's proximity, rather than keeping the entire landscape resident at all times. This drastically lowers the "always-on" memory footprint and prevents the editor from hitching when you perform localized landscape sculpting or painting.
When working with large environments, always verify that your proxy bounds align with your Grid-Based Spatial Loading Cells to ensure efficient data streaming and consistent occlusion culling.
Key Takeaway: Landscape Streaming Proxies transform a heavy, monolithic landscape into modular, streaming-ready chunks, enabling finer control over memory residency and significantly faster editor iteration times.